qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
To: qemu-block@nongnu.org
Cc: fam@euphon.net, kwolf@redhat.com, vsementsov@virtuozzo.com,
	qemu-devel@nongnu.org, mreitz@redhat.com, den@openvz.org,
	jsnow@redhat.com
Subject: [PATCH v5 8/9] block/qcow2-bitmap: fix and improve qcow2_reopen_bitmaps_rw
Date: Fri, 27 Sep 2019 15:23:54 +0300	[thread overview]
Message-ID: <20190927122355.7344-9-vsementsov@virtuozzo.com> (raw)
In-Reply-To: <20190927122355.7344-1-vsementsov@virtuozzo.com>

- Correct check for write access to file child, and in correct place
  (only if we want to write).
- Support reopen rw -> rw (which will be used in following commit),
  for example, !bdrv_dirty_bitmap_readonly() is not a corruption if
  bitmap is marked IN_USE in the image.
- Consider unexpected bitmap as a corruption and check other
  combinations of in-image and in-RAM bitmaps.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
---
 block/qcow2-bitmap.c | 77 +++++++++++++++++++++++++++++++++-----------
 1 file changed, 58 insertions(+), 19 deletions(-)

diff --git a/block/qcow2-bitmap.c b/block/qcow2-bitmap.c
index a636dc50ca..6335d85244 100644
--- a/block/qcow2-bitmap.c
+++ b/block/qcow2-bitmap.c
@@ -1108,18 +1108,14 @@ int qcow2_reopen_bitmaps_rw(BlockDriverState *bs, Error **errp)
     Qcow2BitmapList *bm_list;
     Qcow2Bitmap *bm;
     GSList *ro_dirty_bitmaps = NULL;
-    int ret = 0;
+    int ret = -EINVAL;
+    bool need_header_update = false;
 
     if (s->nb_bitmaps == 0) {
         /* No bitmaps - nothing to do */
         return 0;
     }
 
-    if (!can_write(bs)) {
-        error_setg(errp, "Can't write to the image on reopening bitmaps rw");
-        return -EINVAL;
-    }
-
     bm_list = bitmap_list_load(bs, s->bitmap_directory_offset,
                                s->bitmap_directory_size, errp);
     if (bm_list == NULL) {
@@ -1128,32 +1124,75 @@ int qcow2_reopen_bitmaps_rw(BlockDriverState *bs, Error **errp)
 
     QSIMPLEQ_FOREACH(bm, bm_list, entry) {
         BdrvDirtyBitmap *bitmap = bdrv_find_dirty_bitmap(bs, bm->name);
-        if (bitmap == NULL) {
-            continue;
-        }
 
-        if (!bdrv_dirty_bitmap_readonly(bitmap)) {
-            error_setg(errp, "Bitmap %s was loaded prior to rw-reopen, but was "
-                       "not marked as readonly. This is a bug, something went "
-                       "wrong. All of the bitmaps may be corrupted", bm->name);
-            ret = -EINVAL;
+        if (!bitmap) {
+            error_setg(errp, "Unexpected bitmap '%s' in image '%s'",
+                       bm->name, bs->filename);
             goto out;
         }
 
-        bm->flags |= BME_FLAG_IN_USE;
-        ro_dirty_bitmaps = g_slist_append(ro_dirty_bitmaps, bitmap);
+        if (!(bm->flags & BME_FLAG_IN_USE)) {
+            if (!bdrv_dirty_bitmap_readonly(bitmap)) {
+                error_setg(errp, "Corruption: bitmap '%s' is not marked IN_USE "
+                           "in the image '%s' and not marked readonly in RAM",
+                           bm->name, bs->filename);
+                goto out;
+            }
+            if (bdrv_dirty_bitmap_inconsistent(bitmap)) {
+                error_setg(errp, "Corruption: bitmap '%s' is inconsistent but "
+                           "is not marked IN_USE in the image '%s'", bm->name,
+                           bs->filename);
+                goto out;
+            }
+
+            bm->flags |= BME_FLAG_IN_USE;
+            need_header_update = true;
+        } else {
+            /*
+             * What if flags already has BME_FLAG_IN_USE ?
+             *
+             * 1. if we are reopening RW -> RW it's OK, of course.
+             * 2. if we are reopening RO -> RW:
+             *   2.1 if @bitmap is inconsistent, it's OK. It means that it was
+             *       inconsistent (IN_USE) when we loaded it
+             *   2.2 if @bitmap is not inconsistent. This seems to be impossible
+             *       and implies third party interaction. Let's error-out for
+             *       safety.
+             */
+            if (bdrv_dirty_bitmap_readonly(bitmap) &&
+                !bdrv_dirty_bitmap_inconsistent(bitmap))
+            {
+                error_setg(errp, "Corruption: bitmap '%s' is marked IN_USE "
+                           "in the image '%s' but it is readonly and "
+                           "consistent in RAM",
+                           bm->name, bs->filename);
+                goto out;
+            }
+        }
+
+        if (bdrv_dirty_bitmap_readonly(bitmap)) {
+            ro_dirty_bitmaps = g_slist_append(ro_dirty_bitmaps, bitmap);
+        }
     }
 
-    if (ro_dirty_bitmaps != NULL) {
+    if (need_header_update) {
+        if (!can_write(bs->file->bs) || !(bs->file->perm & BLK_PERM_WRITE)) {
+            error_setg(errp, "Failed to reopen bitmaps rw: no write access "
+                       "the protocol file");
+            goto out;
+        }
+
         /* in_use flags must be updated */
         ret = update_ext_header_and_dir_in_place(bs, bm_list);
         if (ret < 0) {
-            error_setg_errno(errp, -ret, "Can't update bitmap directory");
+            error_setg_errno(errp, -ret, "Cannot update bitmap directory");
             goto out;
         }
-        g_slist_foreach(ro_dirty_bitmaps, set_readonly_helper, false);
     }
 
+    g_slist_foreach(ro_dirty_bitmaps, set_readonly_helper, false);
+    ret = 0;
+
 out:
     g_slist_free(ro_dirty_bitmaps);
     bitmap_list_free(bm_list);
-- 
2.21.0



  parent reply	other threads:[~2019-09-27 14:44 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-27 12:23 [PATCH v5 0/9] qcow2-bitmaps: rewrite reopening logic Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` [PATCH v5 1/9] block: switch reopen queue from QSIMPLEQ to QTAILQ Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` [PATCH v5 2/9] block: reverse order for reopen commits Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` [PATCH v5 3/9] iotests: add test-case to 165 to test reopening qcow2 bitmaps to RW Vladimir Sementsov-Ogievskiy
2019-09-27 20:43   ` John Snow
2019-09-27 12:23 ` [PATCH v5 4/9] block/qcow2-bitmap: get rid of bdrv_has_changed_persistent_bitmaps Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` [PATCH v5 5/9] block/qcow2-bitmap: drop qcow2_reopen_bitmaps_rw_hint() Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` [PATCH v5 6/9] block/qcow2-bitmap: do not remove bitmaps on reopen-ro Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` [PATCH v5 7/9] iotests: add test 260 to check bitmap life after snapshot + commit Vladimir Sementsov-Ogievskiy
2019-09-27 12:23 ` Vladimir Sementsov-Ogievskiy [this message]
2019-09-27 12:23 ` [PATCH v5 9/9] qcow2-bitmap: move bitmap reopen-rw code to qcow2_reopen_commit Vladimir Sementsov-Ogievskiy
2019-09-27 22:41 ` [PATCH v5 0/9] qcow2-bitmaps: rewrite reopening logic John Snow
2019-09-27 23:36 ` John Snow
2019-09-30 10:32   ` Vladimir Sementsov-Ogievskiy

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20190927122355.7344-9-vsementsov@virtuozzo.com \
    --to=vsementsov@virtuozzo.com \
    --cc=den@openvz.org \
    --cc=fam@euphon.net \
    --cc=jsnow@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).